Abstract
BACKGROUND: Frozen section analysis of cyst wall has been advocated as standard of care prior to surgical drainage of pancreatic pseudocysts. Recent studies have indicated that frozen section may be unreliable. Our aim was to investigate the role and accuracy of frozen section analysis of the cyst wall in the operative management of pancreatic cystic lesions. METHODS: A retrospective review was performed of 44 patients who underwent operation for a cystic lesion of the pancreas. RESULTS: Of the 25 patients with a diagnosis of pseudocyst, 9 patients had frozen section intra-operatively. Eight frozen sections demonstrated pseudocyst, and correlated with final pathology. One frozen section demonstrated a true cyst. Twenty-three patients had a final diagnosis of pseudocyst. Nineteen patients had a diagnosis of cystic neoplasm, and six patients had frozen sections. Three frozen sections were consistent with cystic neoplasm and correlated with final pathology. Three frozen sections demonstrated characteristics of pseudocyst, two correlated with final pathology, and one showed cystic neoplasm on final pathology. CONCLUSION: Routine use of frozen section may not be necessary; however, frozen section is useful when the intra-operative appearance does not correlate with pre-operative diagnosis.